Characterization of reflectance variability in the industrial paint application of automotive metallic coatings by using principal component analysis

We have applied principal component analysis to examine trialto-trial variability of reflectances of automotive coatings that contain effect pigments. Reflectance databases were measured from different color batch productions using a multi-angle spectrophotometer. A method to classify the principal components was used based on the eigenvalue spectra. It was found that the eigenvalue spectra follow distinct power laws and depend on the detection angle. The scaling exponent provided an estimation of the correlation between reflectances and it was higher near specular reflection, suggesting a contribution from the deposition of effect pigments. Our findings indicate that principal component analysis can be a useful tool to classify different sources of spectral variability in color engineering. (C) 2013 Society of Photo-Optical Instrumentation Engineers (SPIE).